Re: [Xen-devel] [PATCH 1/3] x86/xen: Introduce a global flag to fix the MSI mask bug

[David Vrabel <david.vrabel@citrix.com>]

On 27/10/14 19:45, Konrad Rzeszutek Wilk wrote:
> On Mon, Oct 27, 2014 at 11:09:42AM +0000, David Vrabel wrote:
>> On 27/10/14 02:44, Yijing Wang wrote:
>>> Commit 0e4ccb1505a9 ("PCI: Add x86_msi.msi_mask_irq() and msix_mask_irq()")
>>> fixed MSI mask bug which may cause kernel crash. But the commit
>>> made MSI code complex. Introduce a new global flag "pci_msi_ignore_mask"
>>> to ignore MSI/MSI-X to fix this issue, it's a cleaner solution.
>>> And the commit 0e4ccb1505a9 will be reverted in the later patch.
>>
>> Reviewed-by: David Vrabel <david.vrabel@citrix.com>
>>
>> In the sense that it keeps the odd Xen behaviour.  But...
>>
>> Konrad, why was this fixed like this in the first place?  IMO, it would
> 
> As 0e4ccb1505a9 explains:
>     PCI: Add x86_msi.msi_mask_irq() and msix_mask_irq()
>     
>     Certain platforms do not allow writes in the MSI-X BARs to setup or tear
>     down vector values.  To combat against the generic code trying to write to
>     that and either silently being ignored or crashing due to the pagetables
>     being marked R/O this patch introduces a platform override.
>     
>     Note that we keep two separate, non-weak, functions default_mask_msi_irqs()
>     and default_mask_msix_irqs() for the behavior of the arch_mask_msi_irqs()
>     and arch_mask_msix_irqs(), as the default behavior is needed by x86 PCI
>     code.
>     
>     For Xen, which does not allow the guest to write to MSI-X tables - as the
>     hypervisor is solely responsible for setting the vector values - we
>     implement two nops.

My question specifically was: could this be fixed by allowing writes to
set/clear the mask bit?  If so, why was this not done and could we do
this now?

David